Some slow Beethoven movements (op. 22, op.31/1...)

[Jo498]

There is a bunch of slow movements in 9/8 and similar (6/8 etc) time signatures where there seems a huge discrepancy among musicians wrt tempo. Briefly the question is: Should one count slow dotted quarters or moderately slow eighth notes?

The sparse historical evidence clearly favors the former, or technically more correct, flowing/fast 8th notes because that's how the tempi were usually given as early metronomes had 50 bpm or so as slowest beat.
Probably because some of these movements like the 2 mentioned in the topic are very ornate, featuring lots of semidemiquavers or faster notes many pianists seem to find slower tempi, i.e. moderate to slow 8th that are so slow (~60-70) that there is hardly any feeling left for the dotted quarters.

AFAIK there are only 2 authentic Beethoven MM markings for x/8 adagios, namely op. 106 that has 92 for 8th, and op.18/1 that has 138! for 8th, both probably better construed as ~30 and 46 for dotted quarter (which Beethoven could not indicate because the MM didn't go down far enough).

They are playable like that but it is quite fast and even disregarding the wide range one could certainly argue that these "appassionato"-Movements would be comparably fast for a slow movement. Still, at leat the op.18 movement is always played so that one has a feeling for the dotted quarters, even if it's 20% slower. The Hammerklavier adagio has of course a range from close to the indicated tempo (13-14 min) to half that (25 min). It's also a very long and complex movement with lots of options for tempo changes and rubato that I don't really want to discuss it; the point is only that is has one of the authentic tempo suggestions for an x/8 adagio.

Now there is one movement that seems so slow that it has to be in 8th, not dotted quarters, namely "largo e mesto" (6/8) from op.10/3. There seems a suggestion by Czerny of 8th = 72 which is much slower than the op.106 (92, although not infrequently played closer to 72!) but still a bit faster than my fastest recording (Gulda at ~66-70, the movement lasts about 8 min instead of more typically 10 min), a typical tempo is around 50-55, i.e. slow, not moderate 8th. But the range here is not as wide as in other pieces. (Another movement that seems so slow that slowish 8th are probably correct is the introduzione in the Waldstein sonata.)

There are tempo suggestions from Czerny and others from the 1830s or so, i.e. pretty close to Beethoven, and there are the sometimes infamous suggestions from Kolisch writing in the 1940s (but they very often are in the same range as Czerny, Moscheles etc)

From what scribbled in my scores the historical suggestion for op.31/1,ii is 116-138 for 8th (or ca. 35-46 for dotted quarters). This is in line with op.18 and much faster than op.106 but in any case clearly for dotted quarters being slow but still perceivable as the pulse, not slow 8th. Of my recordings Gulda is closest to that, even at the upper end (7:55 for the whole ~135 for 8th) then Gould (8:50), also Pollini (which I have not heard). The slowest, Heidsieck is close to 14 min!
Not sure if there is a "typical" tempo but probably around 11 min.
The character does change, faster it appears more parodistic (not implausible, I think) but I actually think the corresponding tempo ~30-33 (90-100 for 8th -> 11-12 min playing time), right below the historical range (and around the op.106 marking of 92) can still keep a reasonable flow and pulse of dotted quarters.

There is an even wider range for the similarly ornate "adagio con molta espressione" in op.22. Gulda is so fast that it seems allegretto so some although it is slower than op.31/1 (as plausible for "con espressione" vs "grazioso"), about 124 for 8th. Pollini is a bit slower (~110-112 which agrees with Czerny's suggestion 100-116, Kolisch goes with the 44-46(132-138) from op.18/1 which would be even faster than Gulda!) On the other end, Gilels needs over 11 min with very slow 8th notes (60-63), almost exactly half of Gulda's tempo.

The difference to op.31 is here that more pianists play it so slowly that there is hardly any feeling for dotted quarters left. While Gilels is extremely slow typical tempi seem around 72-80 for 8th, well below the slowest historical suggestion (either Czerny's or even the 92 from op.106)

As in many similar cases musicians mostly agree about the qualitative ranking, i.e. 31/1,ii faster than 22,ii faster than op.106,iii faster than op.10/3,ii. But the actual tempi can be so different that, I think especially in op.31/1 and 22 the character can change quite a bit.

[prémont]

Thanks for this thorough analysis.

[Jo498]

There might be an "Austrian" tradition to play the op.22 adagio "fast"; Badura-Skoda is the slowest with 8 min., Brendel and Buchbinder are below 7 min, much faster than "average". Or maybe they just follow Czerny

There is another original Beethovenian metronome marking for the 9/8 adagio cantabile from the septet, 8th = 132, so that's another data point, but this movement is not quite as ornate (except for the turn in the theme nothing faster than 16th notes) as the piano sonata movements and very clearly in dotted quarters.

[Holden]

There might be an "Austrian" tradition to play the op.22 adagio "fast"; Badura-Skoda is the slowest with 8 min., Brendel and Buchbinder are below 7 min, much faster than "average". Or maybe they just follow Czerny

There is another original Beethovenian metronome marking for the 9/8 adagio cantabile from the septet, 8th = 132, so that's another data point, but this movement is not quite as ornate (except for the turn in the theme nothing faster than 16th notes) as the piano sonata movements and very clearly in dotted quarters.

The metronome markings are certainly not Beethoven's in Op 22 as he did not own one at that stage of his life so any mention of them is purely speculative.

[Jo498]

As I wrote in the first posting they are from Czerny and Moscheles for op.22.
I mentioned explicitly which ones are directly from Beethoven (op.18/1, 20, 106).

As Czerny was a pupil of Beethoven & his tempo suggestions are mostly in line with Beethoven's own and with e.g. Moscheles and Hummel, other musicians close to Beethoven, they are certainly not "purely speculative" but historically as close as it gets (beyond the composer himself) and plausible evidence.

Obviously, none of this is conclusive in determining a "correct" tempo. But the cumulative is evidence very clearly against slowish 8th notes for such movements but for slow dotted quarters (as there is virtually no evidence for the former but plenty for the latter).

[Holden]

As I wrote in the first posting they are from Czerny and Moscheles for op.22.
I mentioned explicitly which ones are directly from Beethoven (op.18/1, 20, 106).

As Czerny was a pupil of Beethoven & his tempo suggestions are mostly in line with Beethoven's own and with e.g. Moscheles and Hummel, other musicians close to Beethoven, they are certainly not "purely speculative" but historically as close as it gets (beyond the composer himself) and plausible evidence.

Obviously, none of this is conclusive in determining a "correct" tempo. But the cumulative is evidence very clearly against slowish 8th notes for such movements but for slow dotted quarters (as there is virtually no evidence for the former but plenty for the latter).

My apologies, I should have read the post with more care. It's been discussed before but the one sonata that LvB did employ his own metronome markings has been mired in controversy.